Paste the output of db.collection.explain("executionStats") and see it as a readable execution tree, which stages ran, index vs. collection scans, documents examined, and where the time went. Nothing leaves your browser.
Run .explain("executionStats") in the shell or Compass, paste the JSON, and hit Visualize.
Spot COLLSCANs, index usage, docs vs. keys examined, and per-stage timings at a glance.
Your query plan is parsed locally and never uploaded anywhere.
An explain plan is MongoDB's report of how it executed a query, which indexes it used, how many documents and index keys it examined, and how long each stage took. Reading explain output is the single most important skill for diagnosing slow MongoDB queries, and it's how you decide which indexes to add.
Running db.collection.explain("executionStats").find(...) returns the winning query plan plus real execution numbers: nReturned, totalDocsExamined, totalKeysExamined, and per-stage timings. Visualizing it as a tree makes the bottleneck obvious.
A collection scan, MongoDB read every document because no suitable index existed. On large collections this is slow; adding an index that supports the query usually turns it into an IXSCAN.
Look for an IXSCAN stage (with an indexName) rather than a COLLSCAN, and check that totalKeysExamined is close to nReturned.
No, the plan is parsed entirely in your browser and never sent to a server.
VisuaLeaf
VisuaLeaf's Query Profiler runs explain for you and visualizes plans right next to your data, with no copy-paste, across MongoDB and SQL, alongside the visual query builder, schema diagrams, and charts.